Summary of Project 3

Feature extraction, storage, running time (cost)

Submission Time

Dimensions of Extracted Features

Advanced Model

Running Time

Performance

Baseline Prediction Accuracy

Advanced Prediction Accuracy

Advanced models: SVM(1, 2, 4, 9), CNN(3), xgboost(5), gbm(6, 11), majority vote(7, 10), random forest(8), logistic(12)

Fun Images

The prediction of 20 fun images.

Summary of Common Problems

• Ease of Reproducibility(train.R & test.R): Input and Output
• Abosolute path: setwd(“F:/UNI/ADS_2016/project3/output”) Relative path: setwd(“~/output”)